linkparser 1.1.4 → 2.0.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- checksums.yaml.gz.sig +0 -0
- data.tar.gz.sig +0 -0
- data/.simplecov +9 -0
- data/ChangeLog +40 -3
- data/History.md +55 -0
- data/Manifest.txt +6 -4
- data/{README.rdoc → README.md} +56 -53
- data/Rakefile +53 -21
- data/ext/dictionary.c +60 -65
- data/ext/extconf.rb +6 -3
- data/ext/linkage.c +117 -368
- data/ext/linkparser.c +56 -27
- data/ext/linkparser.h +14 -16
- data/ext/parseoptions.c +209 -680
- data/ext/sentence.c +62 -149
- data/lib/linkparser.rb +14 -7
- data/lib/linkparser/dictionary.rb +13 -0
- data/lib/linkparser/linkage.rb +277 -166
- data/lib/linkparser/mixins.rb +2 -2
- data/lib/linkparser/parseoptions.rb +58 -0
- data/lib/linkparser/sentence.rb +21 -34
- data/spec/bugfixes_spec.rb +23 -36
- data/spec/helpers.rb +35 -0
- data/spec/linkparser/dictionary_spec.rb +29 -48
- data/spec/linkparser/linkage_spec.rb +199 -268
- data/spec/linkparser/mixins_spec.rb +9 -24
- data/spec/linkparser/parseoptions_spec.rb +45 -59
- data/spec/linkparser/sentence_spec.rb +36 -56
- data/spec/linkparser_spec.rb +6 -25
- metadata +97 -85
- metadata.gz.sig +0 -0
- data/History.rdoc +0 -41
- data/examples/basic-api.rb +0 -65
- data/examples/readme-example.rb +0 -14
data/ext/extconf.rb
CHANGED
@@ -16,23 +16,26 @@ $CFLAGS << ' ' + `#{pkgconfig} --cflags link-grammar`.chomp
|
|
16
16
|
|
17
17
|
dir_config( 'link-grammar' )
|
18
18
|
|
19
|
-
find_library( "link-grammar", "
|
19
|
+
find_library( "link-grammar", "parse_options_create" ) or
|
20
20
|
abort "Could not find Link-Grammar library" +
|
21
21
|
"(http://www.abisource.com/projects/link-grammar/#download)."
|
22
22
|
|
23
23
|
find_header( "link-grammar/link-includes.h" ) or
|
24
24
|
abort "Could not find link-includes.h"
|
25
25
|
|
26
|
-
|
26
|
+
if have_func( "parse_options_get_screen_width", 'link-grammar/link-includes.h' )
|
27
27
|
$stderr.puts "Your link-grammar library is too old for this binding.",
|
28
28
|
"Please upgrade to the latest version posted here:",
|
29
29
|
" http://www.abisource.com/projects/link-grammar/#download",
|
30
30
|
"and try again."
|
31
|
-
abort "
|
31
|
+
abort "The parse_options_get_screen_width() still defined by the installed link-grammar."
|
32
32
|
end
|
33
33
|
|
34
|
+
have_func( 'dictionary_create' )
|
35
|
+
have_func( 'dictionary_create_lang' )
|
34
36
|
have_func( 'parse_options_get_spell_guess' )
|
35
37
|
have_func( 'linkage_get_disjunct_str' )
|
36
38
|
|
39
|
+
create_header()
|
37
40
|
create_makefile( 'linkparser_ext' )
|
38
41
|
|
data/ext/linkage.c
CHANGED
@@ -1,6 +1,6 @@
|
|
1
1
|
/*
|
2
2
|
* linkage.c - Ruby LinkParser Linkage class
|
3
|
-
* $Id: linkage.c,v
|
3
|
+
* $Id: linkage.c,v 92228378be38 2015/03/02 16:44:04 ged $
|
4
4
|
*
|
5
5
|
* Authors:
|
6
6
|
* * Michael Granger <ged@FaerieMUD.org>
|
@@ -12,17 +12,14 @@
|
|
12
12
|
#include "linkparser.h"
|
13
13
|
|
14
14
|
|
15
|
-
/* --------------------------------------------------
|
16
|
-
* Forward declarations
|
17
|
-
* -------------------------------------------------- */
|
18
|
-
|
19
|
-
static VALUE rlink_linkage_make_cnode_array( CNode * );
|
20
|
-
|
21
|
-
|
22
15
|
/* --------------------------------------------------
|
23
16
|
* Macros and constants
|
24
17
|
* -------------------------------------------------- */
|
25
18
|
|
19
|
+
VALUE display_walls_sym;
|
20
|
+
VALUE display_header_sym;
|
21
|
+
VALUE max_width_sym;
|
22
|
+
|
26
23
|
|
27
24
|
/* --------------------------------------------------
|
28
25
|
* Memory-management functions
|
@@ -31,13 +28,14 @@ static VALUE rlink_linkage_make_cnode_array( CNode * );
|
|
31
28
|
* Allocation function
|
32
29
|
*/
|
33
30
|
static struct rlink_linkage *
|
34
|
-
rlink_linkage_alloc()
|
31
|
+
rlink_linkage_alloc()
|
32
|
+
{
|
35
33
|
struct rlink_linkage *ptr = ALLOC( struct rlink_linkage );
|
36
34
|
|
37
35
|
ptr->linkage = NULL;
|
38
36
|
ptr->sentence = Qnil;
|
39
37
|
|
40
|
-
|
38
|
+
rlink_log( "debug", "Initialized an rlink_LINKAGE <%p>", ptr );
|
41
39
|
return ptr;
|
42
40
|
}
|
43
41
|
|
@@ -46,16 +44,11 @@ rlink_linkage_alloc() {
|
|
46
44
|
* GC Mark function
|
47
45
|
*/
|
48
46
|
static void
|
49
|
-
rlink_linkage_gc_mark( struct rlink_linkage *ptr )
|
50
|
-
|
51
|
-
|
47
|
+
rlink_linkage_gc_mark( struct rlink_linkage *ptr )
|
48
|
+
{
|
52
49
|
if ( ptr ) {
|
53
50
|
rb_gc_mark( ptr->sentence );
|
54
51
|
}
|
55
|
-
|
56
|
-
else {
|
57
|
-
debugMsg(( "Not marking uninitialized rlink_LINKAGE" ));
|
58
|
-
}
|
59
52
|
}
|
60
53
|
|
61
54
|
|
@@ -63,7 +56,8 @@ rlink_linkage_gc_mark( struct rlink_linkage *ptr ) {
|
|
63
56
|
* GC Free function
|
64
57
|
*/
|
65
58
|
static void
|
66
|
-
rlink_linkage_gc_free( struct rlink_linkage *ptr )
|
59
|
+
rlink_linkage_gc_free( struct rlink_linkage *ptr )
|
60
|
+
{
|
67
61
|
if ( ptr ) {
|
68
62
|
linkage_delete( (Linkage)ptr->linkage );
|
69
63
|
ptr->linkage = NULL;
|
@@ -72,10 +66,6 @@ rlink_linkage_gc_free( struct rlink_linkage *ptr ) {
|
|
72
66
|
xfree( ptr );
|
73
67
|
ptr = NULL;
|
74
68
|
}
|
75
|
-
|
76
|
-
else {
|
77
|
-
debugMsg(( "Not freeing an uninitialized rlink_LINKAGE" ));
|
78
|
-
}
|
79
69
|
}
|
80
70
|
|
81
71
|
|
@@ -83,7 +73,8 @@ rlink_linkage_gc_free( struct rlink_linkage *ptr ) {
|
|
83
73
|
* Object validity checker. Returns the data pointer.
|
84
74
|
*/
|
85
75
|
static struct rlink_linkage *
|
86
|
-
check_linkage(
|
76
|
+
check_linkage( VALUE self )
|
77
|
+
{
|
87
78
|
Check_Type( self, T_DATA );
|
88
79
|
|
89
80
|
if ( !IsLinkage(self) ) {
|
@@ -99,7 +90,8 @@ check_linkage( VALUE self ) {
|
|
99
90
|
* Fetch the data pointer and check it for sanity.
|
100
91
|
*/
|
101
92
|
static struct rlink_linkage *
|
102
|
-
get_linkage( VALUE self )
|
93
|
+
get_linkage( VALUE self )
|
94
|
+
{
|
103
95
|
struct rlink_linkage *ptr = check_linkage( self );
|
104
96
|
|
105
97
|
if ( !ptr )
|
@@ -113,7 +105,8 @@ get_linkage( VALUE self ) {
|
|
113
105
|
* Publicly-usable linkage-fetcher
|
114
106
|
*/
|
115
107
|
struct rlink_linkage *
|
116
|
-
rlink_get_linkage( VALUE self )
|
108
|
+
rlink_get_linkage( VALUE self )
|
109
|
+
{
|
117
110
|
return get_linkage( self );
|
118
111
|
}
|
119
112
|
|
@@ -126,8 +119,9 @@ rlink_get_linkage( VALUE self ) {
|
|
126
119
|
* Allocate a new LinkParser::Linkage object.
|
127
120
|
*/
|
128
121
|
static VALUE
|
129
|
-
rlink_linkage_s_alloc( VALUE klass )
|
130
|
-
|
122
|
+
rlink_linkage_s_alloc( VALUE klass )
|
123
|
+
{
|
124
|
+
rlink_log( "debug", "Wrapping an uninitialized Linkage pointer." );
|
131
125
|
return Data_Wrap_Struct( klass, rlink_linkage_gc_mark, rlink_linkage_gc_free, 0 );
|
132
126
|
}
|
133
127
|
|
@@ -187,20 +181,49 @@ rlink_linkage_init( argc, argv, self )
|
|
187
181
|
}
|
188
182
|
|
189
183
|
|
184
|
+
// char * linkage_print_diagram(const Linkage linkage, bool display_walls, size_t screen_width);
|
185
|
+
// void linkage_free_diagram(char * str);
|
186
|
+
//
|
187
|
+
// char * linkage_print_postscript(const Linkage linkage, bool display_walls, bool print_ps_header);
|
188
|
+
// void linkage_free_postscript(char * str);
|
189
|
+
//
|
190
|
+
// char * linkage_print_links_and_domains(const Linkage linkage);
|
191
|
+
// void linkage_free_links_and_domains(char *str);
|
192
|
+
//
|
193
|
+
// char * linkage_print_disjuncts(const Linkage linkage);
|
194
|
+
// void linkage_free_disjuncts(char *str);
|
195
|
+
|
190
196
|
|
191
197
|
/*
|
192
198
|
* call-seq:
|
193
|
-
* diagram -> str
|
199
|
+
* diagram( display_walls: true, max_width: 80 ) -> str
|
194
200
|
*
|
195
|
-
* Return a String containing a diagram of the linkage.
|
201
|
+
* Return a String containing a diagram of the linkage. If +display_walls+ is +true+
|
202
|
+
* the diagram will include the wall-words and connections to them. Strings longer
|
203
|
+
* than +max_width+ will be wrapped at that width.
|
196
204
|
*/
|
197
205
|
static VALUE
|
198
|
-
rlink_linkage_diagram( VALUE self )
|
206
|
+
rlink_linkage_diagram( int argc, VALUE *argv, VALUE self )
|
207
|
+
{
|
199
208
|
struct rlink_linkage *ptr = get_linkage( self );
|
200
209
|
char *diagram_cstr;
|
201
|
-
|
210
|
+
bool display_walls = true;
|
211
|
+
size_t screen_width = 80;
|
212
|
+
VALUE opthash = Qnil,
|
213
|
+
diagram = Qnil;
|
214
|
+
|
215
|
+
rb_scan_args( argc, argv, "0:", &opthash );
|
216
|
+
if ( opthash != Qnil ) {
|
217
|
+
rlink_log_obj( self, "debug", "Got an opthash: %s", RSTRING_PTR(rb_inspect(opthash)) );
|
218
|
+
display_walls = RTEST( rb_hash_lookup2(opthash, display_walls_sym, Qtrue) );
|
219
|
+
if ( rb_hash_lookup(opthash, max_width_sym) != Qnil ) {
|
220
|
+
screen_width = NUM2UINT( rb_hash_lookup(opthash, max_width_sym) );
|
221
|
+
}
|
222
|
+
}
|
223
|
+
|
224
|
+
rlink_log_obj( self, "debug", "Display walls: %d, screen_width: %d", display_walls, screen_width );
|
202
225
|
|
203
|
-
diagram_cstr = linkage_print_diagram( (Linkage)ptr->linkage );
|
226
|
+
diagram_cstr = linkage_print_diagram( (Linkage)ptr->linkage, display_walls, screen_width );
|
204
227
|
diagram = rb_str_new2( diagram_cstr );
|
205
228
|
linkage_free_diagram( diagram_cstr );
|
206
229
|
|
@@ -210,7 +233,7 @@ rlink_linkage_diagram( VALUE self ) {
|
|
210
233
|
|
211
234
|
/*
|
212
235
|
* call-seq:
|
213
|
-
* postscript_diagram(
|
236
|
+
* postscript_diagram( display_walls: true, display_header: false ) -> str
|
214
237
|
*
|
215
238
|
* Returns the macros needed to print out the linkage in a postscript file.
|
216
239
|
* By default, the output is just the set of postscript macros that describe
|
@@ -218,13 +241,26 @@ rlink_linkage_diagram( VALUE self ) {
|
|
218
241
|
* is returned.
|
219
242
|
*/
|
220
243
|
static VALUE
|
221
|
-
rlink_linkage_print_postscript( VALUE
|
244
|
+
rlink_linkage_print_postscript( int argc, VALUE *argv, VALUE self )
|
245
|
+
{
|
222
246
|
struct rlink_linkage *ptr = get_linkage( self );
|
223
247
|
char *diagram_cstr;
|
224
|
-
|
248
|
+
bool display_walls = true,
|
249
|
+
display_header = false;
|
250
|
+
VALUE opthash = Qnil,
|
251
|
+
diagram = Qnil;
|
252
|
+
|
253
|
+
rb_scan_args( argc, argv, "0:", &opthash );
|
254
|
+
if ( opthash != Qnil ) {
|
255
|
+
rlink_log_obj( self, "debug", "Got an opthash: %s", RSTRING_PTR(rb_inspect(opthash)) );
|
256
|
+
display_walls = RTEST( rb_hash_lookup2(opthash, display_walls_sym, Qtrue) );
|
257
|
+
display_header = RTEST( rb_hash_lookup2(opthash, display_header_sym, Qfalse) );
|
258
|
+
}
|
259
|
+
|
260
|
+
rlink_log_obj( self, "debug", "Display walls: %d, display_header: %d", display_walls,
|
261
|
+
display_header );
|
225
262
|
|
226
|
-
diagram_cstr = linkage_print_postscript( (Linkage)ptr->linkage,
|
227
|
-
RTEST(full_doc) ? 1 : 0 );
|
263
|
+
diagram_cstr = linkage_print_postscript( (Linkage)ptr->linkage, display_walls, display_header );
|
228
264
|
diagram = rb_str_new2( diagram_cstr );
|
229
265
|
linkage_free_postscript( diagram_cstr );
|
230
266
|
|
@@ -254,7 +290,8 @@ rlink_linkage_print_postscript( VALUE self, VALUE full_doc ) {
|
|
254
290
|
*
|
255
291
|
*/
|
256
292
|
static VALUE
|
257
|
-
rlink_linkage_links_and_domains( VALUE self )
|
293
|
+
rlink_linkage_links_and_domains( VALUE self )
|
294
|
+
{
|
258
295
|
struct rlink_linkage *ptr = get_linkage( self );
|
259
296
|
char *diagram_cstr;
|
260
297
|
VALUE diagram;
|
@@ -267,72 +304,6 @@ rlink_linkage_links_and_domains( VALUE self ) {
|
|
267
304
|
}
|
268
305
|
|
269
306
|
|
270
|
-
/*
|
271
|
-
* call-seq:
|
272
|
-
* is_fat? -> true or false
|
273
|
-
*
|
274
|
-
* Return +true+ if "fat" linkages were enabled when this linkage was parsed. See
|
275
|
-
* http://www.abiword.org/projects/link-grammar/dict/coordination.html for more
|
276
|
-
* information.
|
277
|
-
*/
|
278
|
-
static VALUE
|
279
|
-
rlink_linkage_is_fat_p( VALUE self ) {
|
280
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
281
|
-
return RTEST( linkage_is_fat(ptr->linkage) ) ? Qtrue : Qfalse;
|
282
|
-
}
|
283
|
-
|
284
|
-
|
285
|
-
|
286
|
-
/*
|
287
|
-
* call-seq:
|
288
|
-
* num_sublinkages -> fixnum
|
289
|
-
*
|
290
|
-
* Return the number of sublinkages for a linkage with conjunctions, 1
|
291
|
-
* otherwise.
|
292
|
-
*/
|
293
|
-
static VALUE
|
294
|
-
rlink_linkage_num_sublinkages( VALUE self ) {
|
295
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
296
|
-
return INT2FIX( linkage_get_num_sublinkages((Linkage)ptr->linkage) );
|
297
|
-
}
|
298
|
-
|
299
|
-
|
300
|
-
/*
|
301
|
-
* call-seq:
|
302
|
-
* current_sublinkage = index -> true or false
|
303
|
-
*
|
304
|
-
* After this call, all operations on the linkage will refer to the index-th
|
305
|
-
* sublinkage. In the case of a linkage without conjunctions, this has no
|
306
|
-
* effect.
|
307
|
-
*/
|
308
|
-
static VALUE
|
309
|
-
rlink_linkage_current_sublinkage_eq( VALUE self, VALUE index ) {
|
310
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
311
|
-
int rval = 0;
|
312
|
-
|
313
|
-
rval = linkage_set_current_sublinkage( (Linkage)ptr->linkage, NUM2INT(index) );
|
314
|
-
|
315
|
-
return INT2FIX( rval );
|
316
|
-
}
|
317
|
-
|
318
|
-
|
319
|
-
/*
|
320
|
-
* call-seq:
|
321
|
-
* current_sublinkage -> fixnum
|
322
|
-
*
|
323
|
-
* Get the index of the current sublinkage.
|
324
|
-
*/
|
325
|
-
static VALUE
|
326
|
-
rlink_linkage_current_sublinkage( VALUE self ) {
|
327
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
328
|
-
int rval = 0;
|
329
|
-
|
330
|
-
rval = linkage_get_current_sublinkage( (Linkage)ptr->linkage );
|
331
|
-
|
332
|
-
return INT2FIX( rval );
|
333
|
-
}
|
334
|
-
|
335
|
-
|
336
307
|
/*
|
337
308
|
* num_words
|
338
309
|
* --
|
@@ -341,7 +312,8 @@ rlink_linkage_current_sublinkage( VALUE self ) {
|
|
341
312
|
* sublinkage.
|
342
313
|
*/
|
343
314
|
static VALUE
|
344
|
-
rlink_linkage_get_num_words( VALUE self )
|
315
|
+
rlink_linkage_get_num_words( VALUE self )
|
316
|
+
{
|
345
317
|
struct rlink_linkage *ptr = get_linkage( self );
|
346
318
|
return INT2FIX( linkage_get_num_words((Linkage)ptr->linkage) );
|
347
319
|
}
|
@@ -353,7 +325,8 @@ rlink_linkage_get_num_words( VALUE self ) {
|
|
353
325
|
* The number of links used in the current sublinkage.
|
354
326
|
*/
|
355
327
|
static VALUE
|
356
|
-
rlink_linkage_get_num_links( VALUE self )
|
328
|
+
rlink_linkage_get_num_links( VALUE self )
|
329
|
+
{
|
357
330
|
struct rlink_linkage *ptr = get_linkage( self );
|
358
331
|
return INT2FIX( linkage_get_num_links((Linkage)ptr->linkage) );
|
359
332
|
}
|
@@ -366,7 +339,8 @@ rlink_linkage_get_num_links( VALUE self ) {
|
|
366
339
|
* current sublinkage.
|
367
340
|
*/
|
368
341
|
static VALUE
|
369
|
-
rlink_linkage_get_link_lword( VALUE self, VALUE index )
|
342
|
+
rlink_linkage_get_link_lword( VALUE self, VALUE index )
|
343
|
+
{
|
370
344
|
struct rlink_linkage *ptr = get_linkage( self );
|
371
345
|
int i = NUM2INT( index );
|
372
346
|
|
@@ -381,7 +355,8 @@ rlink_linkage_get_link_lword( VALUE self, VALUE index ) {
|
|
381
355
|
* current sublinkage.
|
382
356
|
*/
|
383
357
|
static VALUE
|
384
|
-
rlink_linkage_get_link_rword( VALUE self, VALUE index )
|
358
|
+
rlink_linkage_get_link_rword( VALUE self, VALUE index )
|
359
|
+
{
|
385
360
|
struct rlink_linkage *ptr = get_linkage( self );
|
386
361
|
int i = NUM2INT( index );
|
387
362
|
|
@@ -395,7 +370,8 @@ rlink_linkage_get_link_rword( VALUE self, VALUE index ) {
|
|
395
370
|
* The number of words spanned by the index-th link of the current sublinkage.
|
396
371
|
*/
|
397
372
|
static VALUE
|
398
|
-
rlink_linkage_get_link_length( VALUE self, VALUE index )
|
373
|
+
rlink_linkage_get_link_length( VALUE self, VALUE index )
|
374
|
+
{
|
399
375
|
struct rlink_linkage *ptr = get_linkage( self );
|
400
376
|
int i = NUM2INT( index );
|
401
377
|
|
@@ -409,7 +385,8 @@ rlink_linkage_get_link_length( VALUE self, VALUE index ) {
|
|
409
385
|
* The "intersection" of the left and right connectors that comprise the link.
|
410
386
|
*/
|
411
387
|
static VALUE
|
412
|
-
rlink_linkage_get_link_label( VALUE self, VALUE index )
|
388
|
+
rlink_linkage_get_link_label( VALUE self, VALUE index )
|
389
|
+
{
|
413
390
|
struct rlink_linkage *ptr = get_linkage( self );
|
414
391
|
int i = NUM2INT( index );
|
415
392
|
const char *label;
|
@@ -427,7 +404,8 @@ rlink_linkage_get_link_label( VALUE self, VALUE index ) {
|
|
427
404
|
* The label on the left word of the index-th link of the current sublinkage.
|
428
405
|
*/
|
429
406
|
static VALUE
|
430
|
-
rlink_linkage_get_link_llabel( VALUE self, VALUE index )
|
407
|
+
rlink_linkage_get_link_llabel( VALUE self, VALUE index )
|
408
|
+
{
|
431
409
|
struct rlink_linkage *ptr = get_linkage( self );
|
432
410
|
int i = NUM2INT( index );
|
433
411
|
const char *label = NULL;
|
@@ -444,7 +422,8 @@ rlink_linkage_get_link_llabel( VALUE self, VALUE index ) {
|
|
444
422
|
* The label on the right word of the index-th link of the current sublinkage.
|
445
423
|
*/
|
446
424
|
static VALUE
|
447
|
-
rlink_linkage_get_link_rlabel( VALUE self, VALUE index )
|
425
|
+
rlink_linkage_get_link_rlabel( VALUE self, VALUE index )
|
426
|
+
{
|
448
427
|
struct rlink_linkage *ptr = get_linkage( self );
|
449
428
|
int i = NUM2INT( index );
|
450
429
|
const char *label = NULL;
|
@@ -470,10 +449,11 @@ rlink_linkage_get_link_rlabel( VALUE self, VALUE index ) {
|
|
470
449
|
*
|
471
450
|
*/
|
472
451
|
static VALUE
|
473
|
-
rlink_linkage_get_disjunct_strings( VALUE self )
|
452
|
+
rlink_linkage_get_disjunct_strings( VALUE self )
|
453
|
+
{
|
474
454
|
struct rlink_linkage *ptr = get_linkage( self );
|
475
455
|
const char *disjunct;
|
476
|
-
|
456
|
+
unsigned long i, count = 0l;
|
477
457
|
VALUE disjuncts_ary;
|
478
458
|
|
479
459
|
count = linkage_get_num_words( (Linkage)ptr->linkage );
|
@@ -505,7 +485,8 @@ rlink_linkage_get_disjunct_strings( VALUE self ) {
|
|
505
485
|
*
|
506
486
|
*/
|
507
487
|
static VALUE
|
508
|
-
rlink_linkage_get_link_num_domains( VALUE self, VALUE index )
|
488
|
+
rlink_linkage_get_link_num_domains( VALUE self, VALUE index )
|
489
|
+
{
|
509
490
|
struct rlink_linkage *ptr = get_linkage( self );
|
510
491
|
int i = NUM2INT( index );
|
511
492
|
int count = 0;
|
@@ -522,7 +503,8 @@ rlink_linkage_get_link_num_domains( VALUE self, VALUE index ) {
|
|
522
503
|
* Returns the names of the domains the index-th link belongs to.
|
523
504
|
*/
|
524
505
|
static VALUE
|
525
|
-
rlink_linkage_get_link_domain_names( VALUE self, VALUE index )
|
506
|
+
rlink_linkage_get_link_domain_names( VALUE self, VALUE index )
|
507
|
+
{
|
526
508
|
struct rlink_linkage *ptr = get_linkage( self );
|
527
509
|
const char **names;
|
528
510
|
int i = NUM2INT( index );
|
@@ -552,10 +534,11 @@ rlink_linkage_get_link_domain_names( VALUE self, VALUE index ) {
|
|
552
534
|
* The original spellings can be obtained by calls to Sentence#words.
|
553
535
|
*/
|
554
536
|
static VALUE
|
555
|
-
rlink_linkage_get_words( VALUE self )
|
537
|
+
rlink_linkage_get_words( VALUE self )
|
538
|
+
{
|
556
539
|
struct rlink_linkage *ptr = get_linkage( self );
|
557
540
|
const char **words;
|
558
|
-
|
541
|
+
unsigned long count, i;
|
559
542
|
VALUE words_ary;
|
560
543
|
|
561
544
|
count = linkage_get_num_words( (Linkage)ptr->linkage );
|
@@ -570,31 +553,6 @@ rlink_linkage_get_words( VALUE self ) {
|
|
570
553
|
}
|
571
554
|
|
572
555
|
|
573
|
-
/*
|
574
|
-
* call-seq:
|
575
|
-
* compute_union -> true or false
|
576
|
-
*
|
577
|
-
* If the linkage has a conjunction, combine all of the links occurring in all
|
578
|
-
* sublinkages together -- in effect creating a "master" linkage (which may
|
579
|
-
* have crossing links). The union is created as another sublinkage, thus
|
580
|
-
* increasing the number of sublinkages by one, and is returned by this method.
|
581
|
-
* If the linkage has no conjunctions, computing its union has no effect. This
|
582
|
-
* method returns true if computing its union caused another sublinkage to be
|
583
|
-
* created.
|
584
|
-
*/
|
585
|
-
static VALUE
|
586
|
-
rlink_linkage_compute_union( VALUE self ) {
|
587
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
588
|
-
int before, after;
|
589
|
-
|
590
|
-
before = linkage_get_num_sublinkages( (Linkage)ptr->linkage );
|
591
|
-
linkage_compute_union( (Linkage)ptr->linkage );
|
592
|
-
after = linkage_get_num_sublinkages( (Linkage)ptr->linkage );
|
593
|
-
|
594
|
-
return (after > before) ? Qtrue : Qfalse;
|
595
|
-
}
|
596
|
-
|
597
|
-
|
598
556
|
/*
|
599
557
|
* call-seq:
|
600
558
|
* linkage.unused_word_cost -> fixnum
|
@@ -604,7 +562,8 @@ rlink_linkage_compute_union( VALUE self ) {
|
|
604
562
|
*
|
605
563
|
*/
|
606
564
|
static VALUE
|
607
|
-
rlink_linkage_unused_word_cost( VALUE self )
|
565
|
+
rlink_linkage_unused_word_cost( VALUE self )
|
566
|
+
{
|
608
567
|
struct rlink_linkage *ptr = get_linkage( self );
|
609
568
|
int rval;
|
610
569
|
|
@@ -622,7 +581,8 @@ rlink_linkage_unused_word_cost( VALUE self ) {
|
|
622
581
|
*
|
623
582
|
*/
|
624
583
|
static VALUE
|
625
|
-
rlink_linkage_disjunct_cost( VALUE self )
|
584
|
+
rlink_linkage_disjunct_cost( VALUE self )
|
585
|
+
{
|
626
586
|
struct rlink_linkage *ptr = get_linkage( self );
|
627
587
|
int rval;
|
628
588
|
|
@@ -632,25 +592,6 @@ rlink_linkage_disjunct_cost( VALUE self ) {
|
|
632
592
|
}
|
633
593
|
|
634
594
|
|
635
|
-
/*
|
636
|
-
* call-seq:
|
637
|
-
* linkage.and_cost -> fixnum
|
638
|
-
*
|
639
|
-
* Returns the AND cost of the linkage, which is the difference in length
|
640
|
-
* between and-list elements.
|
641
|
-
*
|
642
|
-
*/
|
643
|
-
static VALUE
|
644
|
-
rlink_linkage_and_cost( VALUE self ) {
|
645
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
646
|
-
int rval;
|
647
|
-
|
648
|
-
rval = linkage_and_cost( (Linkage)ptr->linkage );
|
649
|
-
|
650
|
-
return INT2FIX( rval );
|
651
|
-
}
|
652
|
-
|
653
|
-
|
654
595
|
/*
|
655
596
|
* call-seq:
|
656
597
|
* linkage.link_cost -> fixnum
|
@@ -661,7 +602,8 @@ rlink_linkage_and_cost( VALUE self ) {
|
|
661
602
|
*
|
662
603
|
*/
|
663
604
|
static VALUE
|
664
|
-
rlink_linkage_link_cost( VALUE self )
|
605
|
+
rlink_linkage_link_cost( VALUE self )
|
606
|
+
{
|
665
607
|
struct rlink_linkage *ptr = get_linkage( self );
|
666
608
|
int rval;
|
667
609
|
|
@@ -671,65 +613,6 @@ rlink_linkage_link_cost( VALUE self ) {
|
|
671
613
|
}
|
672
614
|
|
673
615
|
|
674
|
-
/*
|
675
|
-
* call-seq:
|
676
|
-
* linkage.canonical? -> true or false
|
677
|
-
*
|
678
|
-
* Returns +true+ if the linkage is canonical. The canonical linkage is the
|
679
|
-
* one in which the minimal disjunct that ever occurrs in a position is used
|
680
|
-
* in that position.
|
681
|
-
*/
|
682
|
-
static VALUE
|
683
|
-
rlink_linkage_canonical_p( VALUE self ) {
|
684
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
685
|
-
int rval = 0;
|
686
|
-
|
687
|
-
rval = linkage_is_canonical( (Linkage)ptr->linkage );
|
688
|
-
|
689
|
-
return rval ? Qtrue : Qfalse;
|
690
|
-
}
|
691
|
-
|
692
|
-
|
693
|
-
/*
|
694
|
-
* call-seq:
|
695
|
-
* linkage.improper? -> true or false
|
696
|
-
*
|
697
|
-
* Returns +true+ if the linkage is "improper".
|
698
|
-
* --
|
699
|
-
* :FIXME: Find out what an "improper fat linkage" is.
|
700
|
-
*
|
701
|
-
*/
|
702
|
-
static VALUE
|
703
|
-
rlink_linkage_improper_p( VALUE self ) {
|
704
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
705
|
-
int rval = 0;
|
706
|
-
|
707
|
-
rval = linkage_is_improper( (Linkage)ptr->linkage );
|
708
|
-
|
709
|
-
return rval ? Qtrue : Qfalse;
|
710
|
-
}
|
711
|
-
|
712
|
-
|
713
|
-
/*
|
714
|
-
* call-seq:
|
715
|
-
* linkage.has_inconsistent_domains? -> true or false
|
716
|
-
*
|
717
|
-
* Returns +true+ if the linkage has inconsistent domains.
|
718
|
-
* --
|
719
|
-
* :FIXME: Find out what it means that a linkage has inconsistent domains.
|
720
|
-
*
|
721
|
-
*/
|
722
|
-
static VALUE
|
723
|
-
rlink_linkage_has_inconsistent_domains_p( VALUE self ) {
|
724
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
725
|
-
int rval = 0;
|
726
|
-
|
727
|
-
rval = linkage_has_inconsistent_domains( (Linkage)ptr->linkage );
|
728
|
-
|
729
|
-
return rval ? Qtrue : Qfalse;
|
730
|
-
}
|
731
|
-
|
732
|
-
|
733
616
|
/*
|
734
617
|
* call-seq:
|
735
618
|
* linkage.violation_name -> str
|
@@ -738,7 +621,8 @@ rlink_linkage_has_inconsistent_domains_p( VALUE self ) {
|
|
738
621
|
* name of the violated rule in the post-process knowledge file.
|
739
622
|
*/
|
740
623
|
static VALUE
|
741
|
-
rlink_linkage_get_violation_name( VALUE self )
|
624
|
+
rlink_linkage_get_violation_name( VALUE self )
|
625
|
+
{
|
742
626
|
struct rlink_linkage *ptr = get_linkage( self );
|
743
627
|
const char *violation_name = NULL;
|
744
628
|
|
@@ -752,133 +636,20 @@ rlink_linkage_get_violation_name( VALUE self ) {
|
|
752
636
|
}
|
753
637
|
|
754
638
|
|
755
|
-
/*
|
756
|
-
* call-seq:
|
757
|
-
* linkage.constituent_tree -> hash
|
758
|
-
*
|
759
|
-
* Return the Linkage's constituent tree as a Array of hierarchical "CTree" structs.
|
760
|
-
*
|
761
|
-
* sent = dict.parse( "He is a big dog." )
|
762
|
-
* link = sent.linkages.first
|
763
|
-
* ctree = link.constituent_tree
|
764
|
-
* # => [#<struct Struct::LinkParserLinkageCTree label="S",
|
765
|
-
* children=[#<struct Struct::LinkParserLinkageCTree label="NP">, ...],
|
766
|
-
* start=0, end=5>]
|
767
|
-
*
|
768
|
-
*/
|
769
|
-
static VALUE
|
770
|
-
rlink_linkage_constituent_tree( VALUE self ) {
|
771
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
772
|
-
CNode *ctree = NULL;
|
773
|
-
VALUE rval = Qnil;
|
774
|
-
|
775
|
-
ctree = linkage_constituent_tree( (Linkage)ptr->linkage );
|
776
|
-
rval = rlink_linkage_make_cnode_array( ctree );
|
777
|
-
|
778
|
-
linkage_free_constituent_tree( ctree );
|
779
|
-
return rval;
|
780
|
-
}
|
781
|
-
|
782
|
-
|
783
|
-
/*
|
784
|
-
* Make an Array of LinkParser::Linkage::CTree objects from the specified
|
785
|
-
* linked list of CNode *.
|
786
|
-
*/
|
787
|
-
static VALUE
|
788
|
-
rlink_linkage_make_cnode_array( CNode *ctree ) {
|
789
|
-
VALUE nodes = rb_ary_new();
|
790
|
-
VALUE rnode;
|
791
|
-
CNode *cnode = ctree;
|
792
|
-
|
793
|
-
/*
|
794
|
-
struct CNode_s {
|
795
|
-
char * label;
|
796
|
-
CNode * child;
|
797
|
-
CNode * next;
|
798
|
-
int start, end;
|
799
|
-
};
|
800
|
-
*/
|
801
|
-
while ( cnode ) {
|
802
|
-
rnode = rb_struct_new( rlink_sLinkageCTree,
|
803
|
-
rb_str_new2( linkage_constituent_node_get_label(cnode) ),
|
804
|
-
Qnil,
|
805
|
-
INT2FIX( linkage_constituent_node_get_start(cnode) ),
|
806
|
-
INT2FIX( linkage_constituent_node_get_end(cnode) ) /* end */
|
807
|
-
);
|
808
|
-
|
809
|
-
/* Make a node array for any children */
|
810
|
-
rb_struct_aset( rnode, INT2FIX(1),
|
811
|
-
rlink_linkage_make_cnode_array(linkage_constituent_node_get_child(cnode)) );
|
812
|
-
|
813
|
-
rb_ary_push( nodes, rnode );
|
814
|
-
cnode = linkage_constituent_node_get_next( cnode );
|
815
|
-
}
|
816
|
-
|
817
|
-
return nodes;
|
818
|
-
}
|
819
|
-
|
820
|
-
|
821
|
-
/*
|
822
|
-
* call-seq:
|
823
|
-
* linkage.constituent_tree_string( mode=1 ) -> str
|
824
|
-
*
|
825
|
-
* Return the constituent tree as a printable string.
|
826
|
-
*
|
827
|
-
* Example:
|
828
|
-
* sent = dict.parse( "He is a big dog." )
|
829
|
-
* link = sent.linkages.first
|
830
|
-
* link.constituent_tree_string
|
831
|
-
*
|
832
|
-
* # ==> "(S (NP He)\n (VP is\n (NP a big dog))\n .)\n"
|
833
|
-
*/
|
834
|
-
static VALUE
|
835
|
-
rlink_linkage_constituent_tree_string( int argc, VALUE *argv, VALUE self ) {
|
836
|
-
struct rlink_linkage *ptr = get_linkage( self );
|
837
|
-
char *ctree_string = NULL;
|
838
|
-
VALUE rval = Qnil, modenum = Qnil;
|
839
|
-
int mode;
|
840
|
-
|
841
|
-
if ( rb_scan_args(argc, argv, "01", &modenum) == 1 ) {
|
842
|
-
mode = NUM2INT( modenum );
|
843
|
-
} else {
|
844
|
-
mode = 1;
|
845
|
-
}
|
846
|
-
|
847
|
-
if ( mode < 1 || mode > 3 )
|
848
|
-
rb_raise( rb_eArgError, "Illegal mode %d specified.", mode );
|
849
|
-
|
850
|
-
ctree_string = linkage_print_constituent_tree( (Linkage)ptr->linkage, mode );
|
851
|
-
|
852
|
-
if ( ctree_string ) {
|
853
|
-
rval = rb_str_new2( ctree_string );
|
854
|
-
linkage_free_constituent_tree_str( ctree_string );
|
855
|
-
} else {
|
856
|
-
rval = Qnil;
|
857
|
-
}
|
858
|
-
|
859
|
-
return rval;
|
860
|
-
}
|
861
|
-
|
862
|
-
|
863
|
-
|
864
|
-
/*
|
865
|
-
* This is the API's representation of a parse. A LinkParser::Sentence may have one or more
|
866
|
-
* of LinkParser::Linkages, each of which represents one possible structure of the sentence.
|
867
|
-
* It can be thought of as a Sentence together with a collection of links. If the parse
|
868
|
-
* has a conjunction, then the Linkage is made up of at least two "sublinkages". A
|
869
|
-
* Linkage can be pretty printed in either ASCII or Postscript format, and individual
|
870
|
-
* links can be extracted.
|
871
|
-
*
|
872
|
-
*/
|
873
639
|
void
|
874
|
-
rlink_init_linkage()
|
640
|
+
rlink_init_linkage()
|
641
|
+
{
|
875
642
|
rlink_cLinkage = rb_define_class_under( rlink_mLinkParser, "Linkage", rb_cObject );
|
876
643
|
|
644
|
+
display_walls_sym = ID2SYM( rb_intern("display_walls") );
|
645
|
+
display_header_sym = ID2SYM( rb_intern("display_header") );
|
646
|
+
max_width_sym = ID2SYM( rb_intern("max_width") );
|
647
|
+
|
877
648
|
rb_define_alloc_func( rlink_cLinkage, rlink_linkage_s_alloc );
|
878
649
|
|
879
650
|
rb_define_method( rlink_cLinkage, "initialize", rlink_linkage_init, -1 );
|
880
|
-
rb_define_method( rlink_cLinkage, "diagram", rlink_linkage_diagram,
|
881
|
-
rb_define_method( rlink_cLinkage, "postscript_diagram", rlink_linkage_print_postscript, 1 );
|
651
|
+
rb_define_method( rlink_cLinkage, "diagram", rlink_linkage_diagram, -1 );
|
652
|
+
rb_define_method( rlink_cLinkage, "postscript_diagram", rlink_linkage_print_postscript, -1 );
|
882
653
|
rb_define_method( rlink_cLinkage, "links_and_domains", rlink_linkage_links_and_domains, 0 );
|
883
654
|
|
884
655
|
rb_define_method( rlink_cLinkage, "num_words", rlink_linkage_get_num_words, 0 );
|
@@ -899,31 +670,9 @@ rlink_init_linkage() {
|
|
899
670
|
rb_define_method( rlink_cLinkage, "words", rlink_linkage_get_words, 0 );
|
900
671
|
rb_define_method( rlink_cLinkage, "disjunct_strings", rlink_linkage_get_disjunct_strings, 0 );
|
901
672
|
|
902
|
-
rb_define_method( rlink_cLinkage, "compute_union", rlink_linkage_compute_union, 0 );
|
903
673
|
rb_define_method( rlink_cLinkage, "unused_word_cost", rlink_linkage_unused_word_cost, 0 );
|
904
674
|
rb_define_method( rlink_cLinkage, "disjunct_cost", rlink_linkage_disjunct_cost, 0 );
|
905
|
-
rb_define_method( rlink_cLinkage, "and_cost", rlink_linkage_and_cost, 0 );
|
906
675
|
rb_define_method( rlink_cLinkage, "link_cost", rlink_linkage_link_cost, 0 );
|
907
|
-
rb_define_method( rlink_cLinkage, "canonical?", rlink_linkage_canonical_p, 0 );
|
908
|
-
rb_define_method( rlink_cLinkage, "improper?", rlink_linkage_improper_p, 0 );
|
909
|
-
rb_define_method( rlink_cLinkage, "has_inconsistent_domains?",
|
910
|
-
rlink_linkage_has_inconsistent_domains_p, 0 );
|
911
676
|
rb_define_method( rlink_cLinkage, "violation_name", rlink_linkage_get_violation_name, 0 );
|
912
|
-
|
913
|
-
/* Struct that contains links of a constituent tree (:label, :children, :start, :end) */
|
914
|
-
rb_define_const( rlink_cLinkage, "CTree", rlink_sLinkageCTree );
|
915
|
-
rlink_sLinkageCTree = rb_struct_define( "LinkParserLinkageCTree",
|
916
|
-
"label", "children", "start", "end", NULL );
|
917
|
-
|
918
|
-
rb_define_method( rlink_cLinkage, "constituent_tree", rlink_linkage_constituent_tree, 0 );
|
919
|
-
rb_define_method( rlink_cLinkage, "constituent_tree_string",
|
920
|
-
rlink_linkage_constituent_tree_string, -1 );
|
921
|
-
|
922
|
-
/* Deprecated sublinkage API */
|
923
|
-
rb_define_method( rlink_cLinkage, "num_sublinkages", rlink_linkage_num_sublinkages, 0 );
|
924
|
-
rb_define_method( rlink_cLinkage, "current_sublinkage=",
|
925
|
-
rlink_linkage_current_sublinkage_eq, 1 );
|
926
|
-
rb_define_method( rlink_cLinkage, "current_sublinkage", rlink_linkage_current_sublinkage, 0 );
|
927
|
-
|
928
677
|
}
|
929
678
|
|